Transparent conductive oxides for perovskite/silicon tandem solar cells by sputter deposition. Model-based characterization of sputter damage, material development and device implementation

Multi-junction solar cells can enable efficiencies beyond the state-of-the-art single junction efficiency limits. Two or more sub-cells make up a multi-junction. Perovskite solar cells, for example, can easily be combined with conventional silicon-based solar cell technologies in monolithically integrated tandem devices. Furthermore, perovskite solar cells offer tremendous potential due to their low material and production costs. The scope of the thesis is the optimization of the transparent front-electrode in perovskite/silicon tandem solar cells and the investigation and reduction of damage during the sputter deposition of the transparent electrode onto the perovskite (sub-)cell. Tin oxide (SnO${}_2$) buffer layers made by thermal atomic layer deposition (ALD) are typically used to shield the perovskite solar cells' sensitive layers from sputter damage. However, this method causes parasitic absorption. We present techniques for sputter damage mitigation that eliminate the need for a SnO${}_2$ buffer layer. Thereby we offer pathways for boosting the tandem device's efficiency further while also simplifying their processing. Several strategies to reduce sputter damage based on established considerations presented in the literature were tested - a low power process, a high-pressure process, and an indirect process. These strategies are in contrast to our standard sputter process, which was initially developed without taking sputter damage into account, concentrating only on the thin film quality. Indium zinc oxide (IZO) served as transparent front-electrode material. Prior to the integration into solar cell devices, we varied the oxygen flow ratio during the sputter processes of the standard process and the various low-damage deposition strategies to optimize the optoelectrical properties of the IZO films. The optimization focused on a good agreement between low parasitic absorption and high electrical conductivity with regard to the application in tandem devices. The low-damage IZO deposition techniques were subsequently tested on semitransparent single-junction perovskite solar cells and compared to the standard IZO deposition. The protective SnO${}_2$-buffer layer was thus removed in order to study how effectively the low-damage techniques reduce sputter damage. One of the studied low-damage techniques, the low power process, exhibited a statistically higher open circuit voltage (V${}_{OC}$) of ~13 mV and a statistically higher fill factor (FF) of ~3 %, compared to the standard IZO deposition process. We then performed light intensity-dependent current density-voltage (J-V) measurements to reveal the correlation between sputter damage and recombination losses. Furthermore, the interactions between different solar cell surface materials and the growing thin film were investigated. The aim was to analyze how the electron transport layer (ETL)/transparent conductive oxide (TCO) interface dynamics and the growth of the TCO are influenced by the choice of the substrate layer on which the TCO is deposited. Therefore, we studied semitransparent perovskite solar cell devices with a low-damage IZO front-electrode deposited on different ETL designs - precisely a C${}_{60}$-only ETL and C${}_{60}$/SnO${}_2$ and C${}_{60}$/PEIE (polyethyleneimine ethoxylated) double layer ETLs. Three key conclusions were drawn from the J-V analysis: 1. when IZO is deposited directly on C${}_{60}$, the resulting J-V curve forms an s-shape; 2. the s-shape formation can be prevented by interlayers, such as SnO${}_2$ or PEIE; 3. the interlayer PEIE leads to an even better performance than the SnO${}_2$ interlayer. The findings were further studied via light intensity-dependent J-V measurements, transient opto-electrical measurements in the all-in-one Paios tool, contact angle analysis, and in-situ grazing incidence small angle x-ray diffraction scattering (GISAXS) measurements during the sputter deposition, monitoring the initial growth behavior of IZO on the various ETL-designs. We found no correlation between the electrical performance and the initial IZO thin film growth. Based on electrical simulations with the SCAPS-1D program, we deduced that the s-shape behavior in C${}_{60}$-only ETL devices results from a potential barrier between the electrode and the ETL. We also found that interfacial non-radiative recombination is not necessarily reflected in the ideality factor. In contrast, interlayers, such as SnO${}_2$ or PEIE, seem to improve charge extraction. Lastly, we transferred our findings into monolithic perovskite/silicon tandem devices. Our goal was to reduce optical losses in tandem devices by removing the protective buffer layer and instead circumvent sputter damage by applying a low-damage IZO deposition process, thereby pushing the overall efficiency and reducing the thermal load. Additionally, this makes the fabrication less time-intense, cheaper, and less complex. Firstly, we performed optical simulations with the MATLAB-based tool GenPro4 of tandem devices with and without a SnO${}_2$-buffer layer to study the optical gain. We observed an overall potential current density gain of 0.6 mA/cm${}^2$ for the sum of both sub-cell currents by removing the SnO${}_2$-buffer layer in the simulation. In the next step, we built monolithic two-terminal tandem devices with a C${}_{60}$/SnO${}_2$ and a C${}_{60}$/PEIE double layer ETL and applied a low-damage IZO deposition process. The superior optics, originating from replacing 20 nm SnO${}_2$ with the ultra-thin ~2 nm PEIE, led to a PCE improvement from 27.4 % to 28.4 % in the tandem devices. The current density loss analysis based on EQE and reflection measurements revealed a gain of 0.58 mA/cm${}^2$ for the sum of the respective sub-cell currents originating from reduced reflection and reduced parasitic absorption. Eventually, we performed long-term stability tests on both tandem device designs to study how the stability of the devices is affected by omitting SnO${}_2$. This work highlights difficulties and offers suitable approaches and implications for depositing the transparent front electrode in ALD SnO${}_2$-buffer layer-free perovskite/silicon tandem systems by industry-relevant means. Additionally, we show the possibility of enhancing the efficiency of tandem solar cells. The results are significant for the development of perovskite/silicon tandem solar cells and will boost the growth of the photovoltaics industry.